Qwen3.5 9B Heretic ARA Refusals5 needs about 7.2 GB of memory at Q4_K_M (a 5.6 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 5 quants, from 5.6 GB (Q4_K_M) to 17.9 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
Qwen3.5 9B Heretic ARA Refusals5 – a 9B model with ARA‑based censorship removal
Qwen3.5 9B Heretic ARA Refusals5 is a 8.95B‑parameter model derived from Qwen3.5‑9B. It uses Heretic v1.2.0’s Arbitrary‑Rank Ablation to reduce refusal rates while maintaining reasoning, tools, and vision abilities. The model is released under apache‑2.0.
Good at
- Reasoning and complex queries
- Tool usage and API integration
- Vision‑enabled tasks
Limits
- No vision support on older hardware
- Large VRAM required for full context
- Draft ARA branch not merged into main
Best for
- Research on censorship mitigation
- Creative writing with reduced refusal
- Testing safety mechanisms
Who it’s for. Researchers and developers interested in exploring ARA‑based model modifications. The model is suitable for academic studies and creative projects that require fewer refusals on English prompts.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 5.6 GB | 7.2 GB | Tight | Get file → | |
| Q5_K_M | 6.5 GB | 8.0 GB | Too big | Get file → | |
| Q6_K | 7.4 GB | 8.9 GB | Too big | Get file → | |
| Q8_0 | 9.5 GB | 11.1 GB | Too big | Get file → | |
| BF16 | 17.9 GB | 19.5 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_M (file size 5.63 GB, estimated 7.2 GB memory at 8 k context). If 12 GB VRAM is available, Q5_K_M (6.47 GB, 8 GB memory) or Q6_K (7.36 GB, 8.9 GB memory) are suitable. For maximum quality, Q8_0 (9.53 GB, 11.1 GB memory) or BF16 (17.92 GB, 19.5 GB memory) can be used on larger GPUs.
What GPU do you need for Qwen3.5 9B Heretic ARA Refusals5?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: the best quant that fits is Q5_K_M (6.5 GB file, about 8.0 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (9.5 GB file, about 11.1 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (9.5 GB file, about 11.1 GB in use).
- 24 GB VRAM: the best quant that fits is BF16 (17.9 GB file, about 19.5 GB in use).
Which Qwen3.5 9B Heretic ARA Refusals5 quant should you choose?
Q4_K_M is the usual starting point for Qwen3.5 9B Heretic ARA Refusals5: it balances file size and quality well. Here it is a 5.6 GB file.
The Q8_0 file is 9.5 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about Qwen3.5 9B Heretic ARA Refusals5
How much VRAM does Qwen3.5 9B Heretic ARA Refusals5 need?
About 7.2 GB at Q4_K_M with an 8K context window. That is the 5.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Qwen3.5 9B Heretic ARA Refusals5 on an 8 GB GPU?
Yes, with Q5_K_M: the 6.5 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen3.5 9B Heretic ARA Refusals5 GGUF?
It depends on the quant: Q4_K_M is 5.6 GB, Q5_K_M is 6.5 GB, Q6_K is 7.4 GB, Q8_0 is 9.5 GB, BF16 is 17.9 GB.
What context length does Qwen3.5 9B Heretic ARA Refusals5 support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.
Where can I download Qwen3.5 9B Heretic ARA Refusals5 GGUF?
The files are on Hugging Face in the NullpoLab/Qwen3.5-9B-Heretic-ARA-Refusals5-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Qwen3.5 9B Heretic ARA Refusals5 use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Qwen3.5 9B Heretic ARA Refusals5 on 8 GB VRAM?
Yes, the Q4_K_M quantized file is 5.63 GB and requires about 7.2 GB VRAM at 8 k context.
Does the model support vision tasks?
The model has vision detection but the current GGUF files do not provide full vision support.
Is the ARA branch stable?
The ARA branch is in draft state as of 2026‑03‑20 and not yet merged into the main Heretic branch.